1. Cross Reference to Related Applications
This patent application is potentially related to the following commonly owned, copending applications:
1. "Method and System for Detecting Optical Faults Within the Optical Domain of a Fiber Communication Network," Ser. No. 08/582,845 by Shoa-Kai Liu, filed on Dec. 28, 1995, and incorporated herein by reference;
2. "System and Method for Photonic Facility and Line Protection Switching," Ser. No. 08/577,663 by John A. Fee, filed on Dec. 22, 1995, and incorporated herein by reference; and
3. "Method and System for Restoration Tributary Switching in a Fiber Network," Ser. No. 08/582,846 by John A. Fee, filed concurrently herewith, and incorporated herein by reference.
2. Field of the Invention
The present invention relates to an optical communication network. More specifically, this invention relates to a system and method for detecting optical faults within a network.
3. Related Art
A communication network serves to transport information among a number of locations. The information is usually presented to the network in the form of time-domain electrical signals and may represent any combination of telephony, video, or computer data in a variety of formats. A typical communication network consists of various physical sites, called nodes, interconnected by information conduits, called "links." Each link serves to carry information from one site to another site. Individual sites contain equipment for combining, separating, transforming, conditioning, and/or routing data.
The traffic of even a single link represents a formidable volume of vital data-equivalent to tens of thousands of phone calls. Sudden failure of a data link can cause a significant loss in revenues for a network owner and loss of commerce and other benefits for the network subscribers. Consequently, restoration techniques have been devised to circumvent a network link failure and to restore normal traffic flow soon.
FIG. 1 shows an example of a typical communications network 100 consisting of sites 101-105 connected by links 120-121. Links are generally implemented using electrical cables, satellites, radio or microwave signals, or optical connections and can stretch for tens or hundreds of miles between sites. Through these links, the communications system 100 carries data signals among the sites 101-105 to effectively interconnect data remote equipments 111-115, i.e. computers, remote terminals, servers, etc. One or more links 120 and 121 that connect two sites are collectively referred to as a span 130.
A span often includes multiple parallel links to increase working and spare capacity. Redundant spare links are commonly added between sites with the intent that they usually carry no data traffic but are available as alternate routes in the event of partial network failure affecting working links. If the network detects a link failure such as a fiber failure, cable cut, or transmitter/receiver nodal failure, traffic may be automatically switched from the failed link to an available spare link.
A typical implementation of a high data rate optical span is depicted in FIG. 2. In FIG. 2, a given site A is connected to another site B by a span consisting of three optical fibers 230,232,234. Two electrical data signals are presented at Site A via inputs 202 and 204. These signals are carried through the network span and recovered at Site B as electrical signal outputs 262 and 264 respectively. For example, these data signals can be STS-48 synchronous data signals each bearing digital data at about 2.5 Gbps or the equivalent of 32 thousand telephone-quality voice channels.
At Site A, the signals enter a digital cross-connect switch 210, and under normal conditions appear as electrical signals along connections 212 and 214. The signal at connection 212 enters an optical Lightwave or Line Terminal Equipment (LTE) 220 shown to include an optical transmitter 221, such as a semiconductor laser. Light emitted by the transmitter 221 is intensity-modulated by the electrical data signal that enters along connection 212 to deliver a modulated optical output signal over optical fiber 230.
After traversing the length of fiber 230, the optical signal arrives at Site B and enters a receiver 241 such as a photodiode. The receiver 241 is shown to be a part of an LTE 230 that amplifies and conditions the signal to render a faithful electrical reproduction at output port 252 of the original electrical data signal provided at input 202. In a similar manner, an electrical data signal presented at input 204 is transported by LTE 222, fiber 232, and LTE 242 to output port 254.
Under normal circumstances, the digital cross-connect switch DCS 260 simply connects port 252 to output port 262 to complete the end-to-end connection of input 202 to output 262. Likewise, DCS 260 normally connects line 254 to output 264 to complete the end-to-end connection of input 204 to output 264.
In FIG. 2, fibers 230 and 232 are referred to as working fibers because they both carry data traffic when all network elements are functioning properly. In contrast, fiber 234 is said to be a spare or "protect" fiber because it carries data traffic only in the event of failure of one of the working fibers 230 or 232 or of the associated LTEs 220, 222, 240, 242. Under normal circumstances, protect fiber 234 does not carry an optical data signal.
When a failure occurs along one of the working fibers 230, 232, digital cross-connect switches 210 and 260 switch data traffic onto the protect fiber 234. For example, if fiber 230 becomes too damaged to transmit light, switch 210 connects input 202 to connection 216. At the same time, DCS 260 disconnects connection 252 and connects output port 262 to connection 256. This switching action restores end-to-end connectivity between input 202 and output 262 despite the failure of working fiber 230.
To successfully perform restoration switching, however, it is necessary to detect failures and to coordinate switching action at each node. As shown in FIG. 2, a separate digital communication network is provided between sites for signaling status and switching commands between DCS 220 and DCS 260. Controller 250 is assigned to Site A to accept alarm inputs 255 from LTE's 220, 222, and 224. Controller 250 also directs the switching action of DCS 210 via control connection 253. A similar Controller 252 resides at Site B to accept alarm inputs 256 from LTEs 240,242, and 244 and to exercise control over DCS 260 via control connection 254. Each Controller 250, 252 is typically an imbedded microprocessor, computer, workstation, or other type of processor for controlling the switching of lightwave terminal equipment, digital cross-connect switches, and optical cross-connect switches.
Controllers 250 and 252 communicate and coordinate with each other over a separate communications link 251. For example, status messages can be sent to indicate, acknowledge, or confirm a link or node state such as an idle, active, inactive, or detected fault state. Any digital signaling protocol can be used such as X.25, Frame Relay, ATM, B-ISDN or Common Channel Signaling 7 protocols. Alternatively, controllers 250 and 251 can communicate status messages using overhead bits or bytes within the data protocol that traverses the working fibers. Restoration algorithms and protocols applied within the controllers to restore end-to-end connectivity in response to a fault detection are well known to those skilled in the art.
Thus, the ability to restore network service depends upon the ability to detect and locate failed network components. Faults have been detected in the electrical domain at LTEs. A transmitter 221 can detect a failed laser diode, for example, by monitoring its bias current. Some transmitters also incorporate a backwave detector. This is a photodiode that picks up the light emerging from the back of the laser diode and can detect the loss of laser output.
A laser diode failure within transmitter 221 can cause LTE 220 to report an alarm condition along an alarm connection 255 to the controller 250. The controller 250 can then direct switching of signals from port 221 to port 216. The controller 250 can also send signals along line 251 to controller 252 at Site B to cause switch 260 to connect port 256 to output port 262.
A receiver 241 can detect the presence of light delivered through fiber 230. A receiver 241 may also monitor bias current of a photodiode, if so equipped. Breakage of fiber 230 or failure of receive element 241 can cause LTE 240 to generate an alarm signal along an alarm connection 256 to controller 252. The controller 252 may then directly control switch 260 and indirectly control switch 210 to circumvent the failed fiber 230 or failed receiver 241.
Another way to detect signal presence at the receive end is by the use of a protocol analyzer that looks for specific data patterns, for example data frame alignment bits, in the digital electrical signal. This detection is usually inherent in the course of handling such signals in the electrical domain. Therefore, LTE 240 could also generate a "loss of frame" alarm to controller 252.
With reference now to FIGS. 3A-3D, the description of operation provided for elements in FIG. 2 applies to the like elements in FIG. 3A-3D. The alarm and controller elements are purposely omitted from the drawing for the sake of clarity.
FIG. 3A is an expansion of FIG. 2 to encompass another span B-C connected to site C. Electrical DCS switches 310,390 are provided at respective endpoint sites A,C. Traffic between sites A and C travels through an intermediate optical/electronic site B. Spans A-B and B-C interconnect the sites as shown. Span A-B includes working fibers 330,332 and protect fiber 334. Span B-C includes working fibers 370,372 and protect fiber 374. Thus, FIG. 3A illustrates the common practice of performing optical-electrical-optical conversion at each site.
In contrast, FIG. 3B represents the more recent, preferred approach of performing more data signal handling in the optical domain. Compared to conventional optical-electronic switching, spans A-B and B-C are optically interconnected at site B. FIG. 3B shows that an entire set of LTEs and a DCS at site B have been replaced by an optical cross connect switch (OCCS) 350 and several optical amplifiers 340-342 and 360-362. Note that the approach of FIG. 3B requires less equipment and removes the intrinsic data rate limitations imposed by intermediate electrical equipment.
The optical switch 350 is capable of being controlled in the same way as the digital cross-connects. An electrical control signal (or an optical control signal) drives the optical switch to selectively form optical connections among various input and output ports facing optical fibers 330-334, 370-374. For example, internal connections 352 and 354 within the OCCS 350 optically couple respective working fibers 330,332,370,372 in both spans A-B, B-C under normal network conditions as shown in FIG. 3B. The optical amplifiers 340-342, 360-362 intensify the lightwave signal and compensate for losses introduced by the working and protect fibers and the OCCS 350.
While the approach of FIG. 3B is advantageous and represents the current trend towards implementing all-optical networks, a new problem is introduced in reliably detecting and locating failures. Without the receivers and LTE's at Site B, it is no longer possible to detect failure along fiber 330 as distinct from failure along fiber 370. In particular, lightwave terminal equipment located at the endpoints of optically interconnected spans cannot distinguish which span causes a communication failure. Therefore, the optical switching function within Site B cannot be usefully employed to bypass failed elements.
This problem is even better understood by examining FIGS. 3C and 3D. In FIG. 3C, fiber 330 has been cut. This failure is detected by receiver 381 at site C. Without a separate detection at intermediate node B, however, the failure cannot be pinpointed to fiber 330 or fiber 370. Any controller receiving an alarm from receiver 381 at an endpoint node must assume that both fibers 330 and 370 have failed.
Restorative switching on a path basis between endpoints can be performed through DCS switches 310 and 390 as shown. For purposes of line protection switching and restoration, this recovery scheme leaves optical switch 350 no more useful than straight permanent connections. This amounts to what is called path protect switching in that entire paths between endpoint sites are switched. The problem is exacerbated as more working and spare fibers are taken into account.
Proper switching of OCCS 350 can provide a greater number of restoration paths. For example, compound failures as depicted in FIG. 3D can be restored. Such versatile restoration line switching can only occur when fault detection capability is provided at Site B to distinguish span A-B failures from span B-C failures.
What is needed is a method and system for detecting optical faults along a fiber network link. Reliable fault detection needs to occur at or near intermediate optical switching sites without impeding data communication traffic. Any detected problems in an all-optical switching network path must be isolated to a specific fiber span or link.